Question about high voltage PCB design

[Ge0rge Marutz]

I am contemplating the design of a printed circuit board to be used as
a heating element. No components will be mounted to the board. I
would just use a long squiggly trace as a heater. The trace would be
terminated at each end of the board with a large solder pad. A wire
will be soldered to these solder pads. The heater will operate up to
400VDC.

What precautions must I take in trace spacing to avoid electrical
arcing? The heating element will not be exposed directly to air. It
will be bare copper that is nickel plated (no solder mask) sandwiched
between a sheet of thermal transfer material and an aluminum block.

How about other nasties such as excessive leakage current?

I plan on using FR-4 board material for this experiment. Do you see
any problems with this?

Could solder fluxes or other residues pose a problem?

I'm asking up front. It's better to be safe than sorry.

Thank you,

Ge0

 

[Phil Allison]

"Ge0rge Marutz"

I am contemplating the design of a printed circuit board to be used as
a heating element. No components will be mounted to the board. I
would just use a long squiggly trace as a heater. The trace would be
terminated at each end of the board with a large solder pad. A wire
will be soldered to these solder pads. The heater will operate up to
400VDC.

** Surely it is easier to spread 100 or so low wattage resistors over a PCB
that your idea ?

A simple series parallel array would have no issues with 400 volts DC.

OTOH - using a SINGLE, whisker fine track that snakes for 50 metres or
more is begging for troubles.




......... Phil

 

[Ge0rge Marutz]

50 meters is rather excessive. I need 11 meters.

"OTOH - using a SINGLE, whisker fine track that snakes for 50 metres
or
more is begging for troubles."

Can you elaborate on this statement?

Thanks,

Ge0rge

 

[Phil Allison]

I am contemplating the design of a printed circuit board to be used as
a heating element. No components will be mounted to the board. I
would just use a long squiggly trace as a heater. The trace would be
terminated at each end of the board with a large solder pad. A wire
will be soldered to these solder pads. The heater will operate up to
400VDC.

** Surely it is easier to spread 100 or so low wattage resistors over a PCB
that your idea ?

A simple series parallel array would have no issues with 400 volts DC.

OTOH - using a SINGLE, whisker fine track that snakes for 50 metres or
more is begging for troubles.

50 meters is rather excessive. I need 11 meters.

** Just how many watts of heat are you generating here ??

11 metres of the finest copper track is not a very high resistance.

"OTOH - using a SINGLE, whisker fine track that snakes for 50 metres
or more is begging for troubles."

Can you elaborate on this statement?

** If it breaks at one tiny spot = no heater.





......... Phil

 

[Ge0rge Marutz]

Trace width can be kept wide enough to where it will not be that
fragile. Just about any board house can reliably manufacture a 1mm
wide trace on 1/2 ounce copper.

I intend on building up a few and put them on long term durability test
to check feasibility. I already have a 20V, 500W version on test. It
has 1300 hours on it so far with no signs of degridation.

My primary concern is with the higher voltage. I typically don't work
with stuff above 20V.

Ge0rge

 

[Phil Allison]

"Ge0rge Marutz" [email protected]> wrote in message

** Please click on "options" and then "reply" so your posts show some
context and name of the other party when they appear on usenet.

Trace width can be kept wide enough to where it will not be that
fragile. Just about any board house can reliably manufacture a 1mm
wide trace on 1/2 ounce copper.

** 11 metres of 1mm wide track is only 10 ohms or so.

Are you for real ??

I intend on building up a few and put them on long term durability test
to check feasibility. I already have a 20V, 500W version on test. It
has 1300 hours on it so far with no signs of degridation.

** That is less than 1 ohm.

At 400 volts you need * 320 ohms * for 500 watts.

Now read my earlier posts again.





........ Phil

 

[nospam]

My primary concern is with the higher voltage.

Your primary concern should be learning what usenet is and how to post on
it, followed by learning ohms law.

--

 

[Noway2]

I am contemplating the design of a printed circuit board to be used as
a heating element. No components will be mounted to the board. I
would just use a long squiggly trace as a heater. The trace would be
terminated at each end of the board with a large solder pad. A wire
will be soldered to these solder pads. The heater will operate up to
400VDC.

One problem I see with this approach is that it doesn't take much heat
to get the copper to delaminate from the FR4.

 

[Joerg]

One problem I see with this approach is that it doesn't take much heat
to get the copper to delaminate from the FR4.

And possibly release some harmful fumes in the process.

Regards, Joerg

 

[Rich Grise]

One problem I see with this approach is that it doesn't take much heat
to get the copper to delaminate from the FR4.

Yeah, well, I've had a bad feeling about it from the start, but I've
"held my tongue" because I don't have the right background; but I'd think
there'd be a way to embed a serpentine piece of nichrome wire in some kind
of mat that would accomplish the same thing.

Good Luck!
Rich

 

[Roy L. Fuchs]

Trace width can be kept wide enough to where it will not be that
fragile. Just about any board house can reliably manufacture a 1mm
wide trace on 1/2 ounce copper.

I intend on building up a few and put them on long term durability test
to check feasibility. I already have a 20V, 500W version on test. It
has 1300 hours on it so far with no signs of degridation.

My primary concern is with the higher voltage. I typically don't work
with stuff above 20V.

Ge0rge

Tell us again what part of "20 volts" is high voltage?

 

[Roy L. Fuchs]

Yeah, well, I've had a bad feeling about it from the start, but I've
"held my tongue" because I don't have the right background; but I'd think
there'd be a way to embed a serpentine piece of nichrome wire in some kind
of mat that would accomplish the same thing.

One is NOT going to dissipate 500 Watts in ANY transducer, whether
it is a piece of trace, wire or other device, and not have a burned
PCB substrate... Every time.

The guy did not give enough info about what his goals are. Not in
the PCB design, but WHAT is the application at hand?

Using a PCB as a heater device mount is not very bright... at all.

What is the job? That IS the question.

 

[Phil Allison]

"Roy L. Fuchs"

Tell us again what part of "20 volts" is high voltage?

** Geezus - what an asinine moron.

**** off - Fuchhead




.......... Phil

 

[redbelly]

Trace width can be kept wide enough to where it will not be that
fragile. Just about any board house can reliably manufacture a 1mm
wide trace on 1/2 ounce copper.

I intend on building up a few and put them on long term durability test
to check feasibility. I already have a 20V, 500W version on test. It
has 1300 hours on it so far with no signs of degridation.

My primary concern is with the higher voltage. I typically don't work
with stuff above 20V.

Ge0rge

George,

There are (at least) a couple of basic problems in what you're
proposing:

If you'd like to use a conducting trace (i.e. low resistance) as the
heater, for practical reasons that means a low voltage & high current
design ... the 400V spec is really not compatible with this.

There was mention of solder flux in your 1st post, so I'll also just
mention that solder is not how heater elements should be connected, if
that is your intention.

Mark

 

[Roy L. Fuchs]

"Roy L. Fuchs"




** Geezus - what an asinine moron.

**** off - Fuchhead

You're a goddamned retard, Allison.

 

[Phil Allison]

"Roy L. Fuchwit-Asshole

You're a goddamned retard, Allison.

** Geezus - what an asinine moron.

**** off - Fuchhead

 

[Roy L. Fuchs]

"Roy L. Fuchwit-Asshole



** Geezus - what an asinine moron.

**** off - Fuchhead

**** off, Philtard...

 

[Zak]

One problem I see with this approach is that it doesn't take much heat
to get the copper to delaminate from the FR4.

I would also wonder about the tolerance of the resistance. Board
probably has a minimume conductivity specified, but not a maximum one.
That said the copper won't be too thick I suppose.

Something similar applies to trace width.


Thomas