rectifier
 

I'll chime in, too, on the opinion that "no caps, and no equalizing
resistors is OK." I've built a lot of full wave center tap diode strings
and full wave full bridge diode strings and with a lot of diodes. I
_would_ recommend, however, that you put in lots of diodes (eg. all 1 kv).
I put in enough in each leg so that my PRV sum for the diodes is *at
least* two times the peak AC volts. So, for a 1 kv DC output, I would have
at least a sum of 3 kv of PRV diodes (or, three 1 kv diodes in series,
preferably four, for each leg). If you get spikes going up that high, they
are usually going to be very transient and the diode will act like a zener
diode and start drawing zener current in the reverse direction. As soon as
that happens, it puts a _load_ on the transient and, unless there is a ton of
energy in the transient and it is quite long, this will keep the voltage
from going higher. I've seen lots of gear with lots of diodes in the HV
legs without balancing resistors or capacitors, and I've been using big
diode strings in 2-3 KV DC supplies (kw level) for decades, and no
problems. Also, the inductance in the primary of big transformers should
tend to limit current surges anyway. You might get bigger voltage surges
when you turn the power switch off.

===== no change to below, included for reference and context =====

On Thu, 12 Apr 2007, Michael Black wrote:

GregS ) writes:
In article , "P.Gregory" wrote:
Wise?
I could put just as many links saying the opposite , as in series , Not
parralel and not paraletic , few few ohms aint gunna do what you yanks say
"diddly squat" unless you get perfect diodes , read as matched ..paul



"Bryan" wrote in message
...
It would be wise to NOT use load-sharing (aka equalizer) resistors &
capacitors. With modern rectifiers, if more than one diode is in a
string,
there will be no problem as long as they are of the same type number. See
http://www.somis.org/D-amplifiers3.html.
73,
Bryan WA7PRC

P.Gregory wrote:
Also do not forget load sharing resistors and bypass caps...
"GregS" wrote in message
...
I am having some difficulty finding a bridge rectifier of 3 Kv and 3
amps.

??

thanks
greg
N6GS

I think I am willing to try using them without caps and resistors.
I allready have 1KV 3 A HER308's. And I have a bunch of thermal
epoxy. I just might purchase some of these to make things easier.

http://www.fairchildsemi.com/ds/FF%2FFFPF04F150S.pdf

greg

The parallel resistors and capacitors in parallel with series diodes
dates from the early days of semiconductor diodes.

The resistors were there to "equalize the reverse voltage drops".
The diodes back then weren't particularly great, and there were worries,
I can't remember if real or imagined, that the diodes couldn't handle
the reverse voltage they'd see. The diodes were there to make sure each
saw the same reverse voltage.

The parallel capacitors were to protect against spikes. Again at a time
when there was real or imagined worry that the precious semiconductor
diodes would be easily damaged.

This was the same period that would always warn about heatsinking the
leads of semiconductors when soldering them, to protect them against
heat.

YOu'll see this talk of parallel resistors and capacitors with the
use of series diodes in older handbooks and magazines, and then it fades
away. I doubt it will hurt anything, but either the diodes got better (in
those early days one often had no choice but to put some diodes in series
just to handle average B+ voltages, let alone outright high voltage), or
the whole thing was an illusion, and wasn't actually needed.

Series resistors limit the current, to protect the diodes against current
surges. The 1971 ARRL Handbook says you need them if the winding resistance
of the transformer are too low to do the limiting itself.

Michael VE2BVW

rectifier
 

------A total SNIP of everything and on to a better idea.-------


Consider using avalanche rectifiers and it makes all the stuff about the
caps and resistors a mute point. There are some good application notes
out there on this subject and the whys and wherefores of it all.

For your needs the BYM36E might be a good choice. At around 30 cents a
piece hard to beat.

http://www.alliedelec.com/Search/Sea...61EC8009E6617F

Best of luck in your project!

Regards,
Paul


GregS wrote:
I am having some difficulty finding a bridge rectifier of 3 Kv and 3 amps.

??

thanks
greg
N6GS

rectifier
 

Paul E. Cater wrote:
------A total SNIP of everything and on to a better idea.-------


Consider using avalanche rectifiers and it makes all the stuff about the
caps and resistors a mute point. There are some good application notes
out there on this subject and the whys and wherefores of it all.

For your needs the BYM36E might be a good choice. At around 30 cents a
piece hard to beat.


http://www.alliedelec.com/Search/Sea...61EC8009E6617F

Best of luck in your project!

Regards,
Paul


GregS wrote:
I am having some difficulty finding a bridge rectifier of 3 Kv and 3
amps.

??

thanks
greg
N6GS

I perused the datasheets of both the BYM36E and 1N5408. Of the
specifications that could matter (that differed):

Reverse Current:
1N5408 = 5uA @ 25°C Ta, 500uA @ 100°C Ta
BYM36E = 5uA @ 25°C Tj, 100uA @ 150°C Tj

Reverse Recovery Time:
1N5408 = Not Rated
BYM36E = 150ns

Nonrepetitive Peak Forward Surge Current:
1N5408 = 200A
BYM36E = 65A

While the BYM36E shows better Reverse Current figures at elevated
temperature (and the measurement point isn't the same), I think it unlikely
that the OP is going to operate them at such temperatures. Considering the
application is 50/60Hz (Period = 20/16.6 ms), Reverse Recovery Time is not a
factor. However, when it comes to Nonrepetitive Peak Forward Surge Current,
the 1N5408 is clearly better. Allied shows 23 hits for 1N5408, the least
expensive being $0.092 each in 1-24 piece lots (stk #568-0554).

Bryan

rectifier
 

Bryan wrote:
Paul E. Cater wrote:
------A total SNIP of everything and on to a better idea.-------


Consider using avalanche rectifiers and it makes all the stuff about the
caps and resistors a mute point. There are some good application notes
out there on this subject and the whys and wherefores of it all.

For your needs the BYM36E might be a good choice. At around 30 cents a
piece hard to beat.


http://www.alliedelec.com/Search/Sea...61EC8009E6617F
Best of luck in your project!

Regards,
Paul


GregS wrote:
I am having some difficulty finding a bridge rectifier of 3 Kv and 3
amps.
??

thanks
greg
N6GS

I perused the datasheets of both the BYM36E and 1N5408. Of the
specifications that could matter (that differed):

Reverse Current:
1N5408 = 5uA @ 25°C Ta, 500uA @ 100°C Ta
BYM36E = 5uA @ 25°C Tj, 100uA @ 150°C Tj

Reverse Recovery Time:
1N5408 = Not Rated
BYM36E = 150ns

Nonrepetitive Peak Forward Surge Current:
1N5408 = 200A
BYM36E = 65A

While the BYM36E shows better Reverse Current figures at elevated
temperature (and the measurement point isn't the same), I think it unlikely
that the OP is going to operate them at such temperatures. Considering the
application is 50/60Hz (Period = 20/16.6 ms), Reverse Recovery Time is not a
factor. However, when it comes to Nonrepetitive Peak Forward Surge Current,
the 1N5408 is clearly better. Allied shows 23 hits for 1N5408, the least
expensive being $0.092 each in 1-24 piece lots (stk #568-0554).

Bryan



Good day all,

There are other factors that make avalanche rectifiers a good choice
over standard types. Of course he should not exceed the specifications
on the data sheets for either. Comparing apples to oranges here and
overlooking the advantages of the basic behaviors of an avalanche device
in an application that can be plagued with spikes.

Here is a couple references on the subject. There are many more out and
about.

http://www.dynexsemi.com/assets/DNX_AN5370_Jul02.pdf

http://www.highvoltageconnection.com...che-diodes.htm


The "non-snappy" behavior, minimizes circuit ringing and can reduce RFI
too! They are rugged and able to withstand reverse energy much better
than conventional diodes.

If the OP was actually thinking of using resistors and caps as has been
batted around relentlessly in this group and others, and when compared
to other avalanche class devices, the .30 cents is a good price.

I have no doubt that if he decides to slap in 5408's with or without all
the extra stuff, it will give him years of service taking for granted he
has the PIV and "I" requirements properly. I simply wanted to offer him
an alternative that eliminates the compensation network argument and
adds another layer of protection.

Best regards to all,
Paul
WD8OSU









It is up to him as to how he wants to approach and remedy his situation.
He is the engineer on the project.

rectifier
 

In article , "Bryan" wrote:
GregS wrote:
P.Gregory wrote:
Wise?
I could put just as many links saying the opposite , as in series , Not
parralel and not paraletic , few few ohms aint gunna do what you yanks
say
"diddly squat" unless you get perfect diodes , read as matched ..paul



Bryan wrote:
It would be wise to NOT use load-sharing (aka equalizer) resistors &
capacitors. With modern rectifiers, if more than one diode is in a
string,
there will be no problem as long as they are of the same type number.
See
http://www.somis.org/D-amplifiers3.html.
73,
Bryan WA7PRC

P.Gregory wrote:
Also do not forget load sharing resistors and bypass caps...
GregS wrote:
I am having some difficulty finding a bridge rectifier of 3 Kv and 3
amps.

??

thanks
greg
N6GS

I think I am willing to try using them without caps and resistors.
I allready have 1KV 3 A HER308's. And I have a bunch of thermal
epoxy. I just might purchase some of these to make things easier.

http://www.fairchildsemi.com/ds/FF%2FFFPF04F150S.pdf

greg

The HER308 looks like a 1N5408 with a specified reverse recovery time. I
like specifications those FFPF04F150S damper diodes. Digi-Key has them
available in only large quantities (non-stocked) but Mouser shows them
in-stock in small lots. Not knowing the dielectric capability of the plastic
TO-220 case, I'd probably mount them (with heatsinks) to an insulating base.

i was trying to get an order together and ordered all parts from Digikey. I got into
discussion with support, and I said the catalog could be a lot smaller and easier
to hold if they listed stocked parts. From the very beginning I always had problems
with DigiKey with stocked parts. Ordered the diodes from Mouser. Nice that many companies
have no minimum orders. That was a big deal years ago.

greg

rectifier
 

GregS wrote:
i was trying to get an order together and ordered all parts from Digikey. I got into
discussion with support, and I said the catalog could be a lot smaller and easier
to hold if they listed stocked parts. From the very beginning I always had problems
with DigiKey with stocked parts. Ordered the diodes from Mouser. Nice that many companies
have no minimum orders. That was a big deal years ago.

greg

Hi Greg,
Just to satisfy my curiosity, what is this power supply for? Mouser &
DigiKey have pretty good reputations. My company orders a lot from
both (for prototypes). I prefer to use their online catalogs, because
they show available stock -- you can't do that with a paper catalog.
73,
Bryan WA7PRC