High Voltage Diode question

I have to rectify a signal. I have a transformer with appx. 4kVAC @
10mA output (unknown frequency). I'm looking for output anywhere from
4-6kVDC. It's not a precise situation, I'm sure the load will work
down to 3kVDC and as high as 6kVDC. So, I need a diode rated for at
least 4kV, preferably 6kV. Acquiring the filtering caps isn't a
problem. I know I could just order one like the VG22 on this page
http://www.amazing1.com/parts.htm
but I figured I'd ask first. What would you guys recommend for a
situation like this? The current should never exceed 25mA, and the
voltage should never exceed 6kV. Any favorite diodes out there? What
configuration should I use? TIA, Steve

 

[Ghost Chip]

Be aware that if you use a single rectifier diode in half wave style, then a
capacitor, the capacitor will charge to the ac peak voltage on one ac peak
(about 5,600 volts. Then on the next opposite half cycle the ac goes
to -5,600 volts. This puts 11,200 volts across the diode. You need one
rated 15,000 volts. You can put diodes is series if you use high value
resistors to divide the voltage to insure that any diode doesn't get more
than its rating.
Zoram

<its you> wrote in message
news:vt8ekv400iqc97c41u8itgo814o373k1o0@4ax.com...

I have to rectify a signal. I have a transformer with appx. 4kVAC @
10mA output (unknown frequency). I'm looking for output anywhere from
4-6kVDC. It's not a precise situation, I'm sure the load will work
down to 3kVDC and as high as 6kVDC. So, I need a diode rated for at
least 4kV, preferably 6kV. Acquiring the filtering caps isn't a
problem. I know I could just order one like the VG22 on this page
http://www.amazing1.com/parts.htm
but I figured I'd ask first. What would you guys recommend for a
situation like this? The current should never exceed 25mA, and the
voltage should never exceed 6kV. Any favorite diodes out there? What
configuration should I use? TIA, Steve

 

Well, I was planning on using four in a full wave rectifier. I've
forgotten the formulas for calculating voltages in each individual
diode for full wave rectifiers, and now I'm second guessing myself.
Am I right to say that for for a full wave rectifier, each half cycle
a pair of diodes will see the full voltage, ~5.6kV peak, like you
said, so each diode in the active pair will have a 2.8kV drop for a
half cycle? Which voltages would you recommend for a full wave
rectifier, still 15kV? My local supplier is out of 12kV 10mA
rectifiers, the only HV ones they have, but those would be just about
perfect for this job I believe, as they aren't expensive and there's
no shipping charge. The transformer is rated at 10mA though, so there
would be no margin for error and they'd probably get pretty hot, but
that's ok with me. I would order a few of the NTE518's like Watt Sun
mentioned, but I don't really want to spend 7 or 8$ in shipping for
4-8 diodes (call me cheap, it's ok). So, in conclusion to my babble,
which voltage range would you recommend for full wave rectification of
4kV signal? Thanks for the response, Steve
On Sat, 23 Aug 2003 17:03:31 -0700, "Ghost Chip" <zoram@cox.net>
wrote:

Be aware that if you use a single rectifier diode in half wave style, then a
capacitor, the capacitor will charge to the ac peak voltage on one ac peak
(about 5,600 volts. Then on the next opposite half cycle the ac goes
to -5,600 volts. This puts 11,200 volts across the diode. You need one
rated 15,000 volts. You can put diodes is series if you use high value
resistors to divide the voltage to insure that any diode doesn't get more
than its rating.
Zoram

its you> wrote in message
news:vt8ekv400iqc97c41u8itgo814o373k1o0@4ax.com...
I have to rectify a signal. I have a transformer with appx. 4kVAC @
10mA output (unknown frequency). I'm looking for output anywhere from
4-6kVDC. It's not a precise situation, I'm sure the load will work
down to 3kVDC and as high as 6kVDC. So, I need a diode rated for at
least 4kV, preferably 6kV. Acquiring the filtering caps isn't a
problem. I know I could just order one like the VG22 on this page
http://www.amazing1.com/parts.htm
but I figured I'd ask first. What would you guys recommend for a
situation like this? The current should never exceed 25mA, and the
voltage should never exceed 6kV. Any favorite diodes out there? What
configuration should I use? TIA, Steve

 

[Ghost Chip]

In the full wave bridge arrangement, each diode sees 1.414 times the ACrms
voltage, i.e. the peak voltage which in your case is ~5.6kv. The two
non-conducting diodes see that as the reverse voltage while the other two
are conducting. You also need to worry about the max surge current a diode
sees. Assume the power is turned on exactly at the AC peak point and the
downstream capacitor is not charged so it looks like a short to ground for
the brief moment. Current for this "pulse" can be limited by the dc ohms of
the secondary winding. E=I*R or a resistor in series with the winding. The
diodes should be able to handle 1.5 to 2 times this expected current pulse
for a few milliseconds.
Good luck. Zoram

 

[Fred Abse]

On Sun, 24 Aug 2003 01:03:31 +0100, Ghost Chip wrote:

Be aware that if you use a single rectifier diode in half wave style,
then a capacitor, the capacitor will charge to the ac peak voltage on
one ac peak (about 5,600 volts. Then on the next opposite half cycle
the ac goes to -5,600 volts. This puts 11,200 volts across the diode.
You need one rated 15,000 volts. You can put diodes is series if you
use high value resistors to divide the voltage to insure that any diode
doesn't get more than its rating.
Zoram

This applies equally to full wave bridges, too.

--
Then there's duct tape ...
(Garrison Keillor)
nofr@sbhevre.pbzchyvax.pb.hx

 

[Fred Abse]

On Sun, 24 Aug 2003 18:38:43 +0100, Fred Abse wrote:

This applies equally to full wave bridges, too.

s/bridge/bi-phase

Realized what I'd said the moment I posted it ;-(

--
Then there's duct tape ...
(Garrison Keillor)
nofr@sbhevre.pbzchyvax.pb.hx