Chris,
Most MOVs are rated as 'Y' components. The manufacturer should be able to
point you towards the appropriate agency information. You should also be
concerned about the overvoltage affects of MOVs. Without some sort of
thermal protection these devices can be very destructive. But, you need to
be careful as the thermal devices can reduce the ability of the MOV to
handle the necessary surge current. If your testing to EN61000-4-5, I have
used the G4A and G5A thermal cut-outs from Thermodisk in series with the
MOVs to reduce the affects of an overvoltage MOV.
I know that there are limitations at placing MOVs to ground in some of the
European countries. To get by you can place an MOV in series with a gas
discharge tube (GDT). This reduces the amount of leakage current as the MOV
(2 nf) is in series with the GDT (1 pf) resulting in a net impedance closer
to the 1 pf of the GDT. The only side affect of this is a high voltage
spike that results from the firing voltage of the GDT. This can be reduced
with a LC filter after the MOV/GDT. This will require some tuning, but
results in good protection that is capable of being used in multiple
countries.
There are some high-current diodes available from Sussex Semiconductor
(www.sussexsemiconductor.com). The series is the AX4. These components are
good up to 6 kA of 8/20 us transient current. I have used them in numerous
products and they work well. They are expensive, but allot less cost than a
series parallel string that is capable of the same performance.
This probably more info than you really want, but if you want more
information or discuss off-line, please let me know.
Thanks,
Bryan Cole
Director of Engineering
Product Safety Officer
Control Concepts / Liebert
Binghamton, NY
T: 607.724.1352 extension 238
M: 607.624.8299
F: 607.724.0153
E: bryan.c...@control-concepts.com
From: Chris Maxwell [mailto:chris.maxw...@nettest.com]
Sent: Friday, May 02, 2003 3:51 PM
To: EMC-PSTC Internet Forum
Subject: Vicor Power Supply Surge Immunity--- More testing...
All,
It seems that I could possibly pass the testing with a suppression device
between Line and Chassis (probably from Neutral to chassis as well).
Some experiments with TVS diodes that I have on hand have been promising.
However, the diodes can't stand up to the multiple surges. i.e. I get
passing test results until the TVS diodes are reduced to a broken pile piece
of Silicon with all of its smoke let out.
I'm also concerned with leakage current.
My design already has a MOV from Line to Neutral; but I have heard that such
components are taboo from Line to Chassis or Neutral to Chassis.
Is there a Y rated TVS, MOV, Sidactor... out there?
I'm tempted to try some 4700pF Y caps that I have around to see if they'll
soak up the surge a little bit.
Chris Maxwell | Design Engineer - Optical Division
email chris.maxw...@nettest.com | dir +1 315 266 5128 | fax +1 315 797 8024
NetTest | 6 Rhoads Drive, Utica, NY 13502 | USA
web www.nettest.com | tel +1 315 797 4449 |
-Original Message-
From: Chris Maxwell
Sent: Friday, May 02, 2003 8:33 AM
To: 'EMC-PSTC Internet Forum'
Subject: Vicor Power Supply Surge Immunity
Colleagues,
Has anyone out there designed power supplies based on the Vicor 2nd
generation converters? I have a design based on the Vicor FARM (Filter
And Rectification Module) with two Vicor V300 series DC-DC converters behind
it.
I can't get this design through surge testing at 500V/250V, even though
the data sheets for all of the modules claim immunity to 2KV/1KV. I'm not
taking issue with the data sheets, Vicor makes good stuff. I just can't
figure out what's up with this design.
I have followed the spec sheet protection recommendations to the letter.
I have suspected the Corcom 6ED8 filter in front could be causing ringing
of the surge; but found that the failure occurs with or without the filter.
The failure is very strange in that it:
-Is load dependent. It seems that a higher load makes failure more likely
-It happens at a random time, up to 17seconds after the surge.
-Is characterized by the FARM's enable signal (which controls the V300s
behind it) going low.
-I can get some improvment (pass 500/250 but fail 1000/500) if I remove
chassis ground from the FARM heatsink.
It almost seems as though the surge is passing through the FARM to its
grounded heatsink, along the way, maybe it is heating up some components,
perhaps this heat takes a little while to get to the FARM's thermal shutdown
circuit, which then dumps the enable signal (I know, really far fetched.)
Maybe a layout problem?
I'd appreciate any experience that others could share with this.
Thanks,
Chris Maxwell | Design Engineer - Optical Division
email chris.maxw...@nettest.com | dir +1 315 266 5128 | fax +1 315 797
8024
NetTest | 6 Rhoads Drive, Utica, NY 13502 | USA
web www.nettest.com | tel +1 315 797 4449 |
This message
