Hi Rich,
The K2AW modules are rated 1.0A at 55 degrees C, derating to zero at 150
degrees C. Thus, a bridge made of four such modules would have a 2.0A/55
deg C rating, attached to any surface having a temperature equal to or below
55 degrees C. No heatsinking required for this rating, just any mounting
surface. I have found, using an ~ 2" x 5" extruded aluminum heat sink,
using one module, I can run it all day at 3 Adc rectified current. The
diodes used internally are 3A diodes, PIV = 1000V, Vf ~ 0.9V @ 3A, Ir < 10uA
@ PIV, with 6, 8, 10 or 14 diodes in series. Thus, the 10kV stack
dissipates ~10W, including forward and reverse losses. Not very much.
There is no argument that potted modules take up more space than diodes
mounted on a PCB. However, the potted module is insulated to its BV rating,
and thus, except for the terminals themselves, may be mounted without
special attention to much of anything. A board would have to be mounted on
the filter cap terminals, or ceramic standoffs, or something to float
(isolate) them.
Not that any of this stuff is high-tech...10kV is not high voltage to me. I
designed power supplies for X-Ray systems (250kV). That's high voltage.
WB2WIK/6
"Each success only buys an admission ticket to a more difficult problem." --
Henry Kissinger
> -----Original Message-----
> From: 2 [SMTP:2@vc.net]
> Sent: Monday, November 19, 2001 2:56 PM
> To: Steve Katz; 'Jim Barber'; AMPS
> Subject: RE: [AMPS] 8877 rectifier board considerations
>
> >
> >
> >> Hi Rich,
> >>
> >>
> >> // We sell 1n5408, 1000piv, 200A peak, 3A avg. diodes for 17-cents
> each
> >> or $14.40 per hundred. . . When mounted on perfboard, the junctions
> run
> >> cooler than epoxy-potted diode-blocks because cooling air has better
> >> access.
> >>
> >> ::Price is right, for sure; although the JTX1N5554 NOS surplus mil-spec
> >> devices, which are more modern design and have about 1/10th the reverse
> >> leakage current of 1N5408's (due to junction design, which is epitaxial
> >> vs. double diffused), are also around $.17/ea or so. I buy them for
> that
> >> price from Apex. They're factory-sealed in serialized containers,
> >> although as with any "old stock," one is never totally sure of
> >> availability!
> >>
> >> Re the cooling issue, though: As you know, axial lead rectifiers are
> >> cooled via conduction, via their leads. Cooling the diode "bodies"
> does
> >> essentially nothing, the thermal resistance from junction to air is
> >> extremely high. The thermal resistance from junction to lead, however,
> is
> >> very low. The leads must be cooled. Flowing air across the leads
> isn't
> >> an effective way to cool diodes because the leads have such little
> surface
> >> area (.040" diameter, typically), and if the leads are "short," this
> >> situation is even worse -- now, there's essentially _no_ surface area.
> >>
> >> So, the modules are actually very well cooled because they are die
> >> cast in a thermoset (thermally conductive) epoxy compound ...
>
> So what is the avg current rating of the same 1n5408/3A diodes when they
> are potted by K2AW? I have used potted diodes and - with the requisite
> heatsink - they were larger than the low-tech perfboard version.
>
> cheers, Steve.
>
> - R. L. Measures, 805.386.3734,AG6K, www.vcnet.com/measures.
> end
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|