TCB10 Failure was (QO Breakers and Panels) [RE-wrenches]

Tue Mar 20 14:00:54 PST 2001

Dear Travis:

Sorry I did not reply earlier but I was on vacation in the Alaskan interior
and didn't have email.

As I mentioned to you during our previous conversations, I agree that the
combiner box failed due to an arc forming on the back of the board.
The pictures of the damage you sent to me in Email looks to be consistent
with arcing.
Once the arc forms it doesn't take anywhere near that much voltage to
sustain it so, as was pointed in other posts, it will just keep on going
until it runs out of fuel.

The maximum voltage rating of the unit is 125 Vdc under all conditions.
The board was designed using UL's guidelines for PCB spacings in a NEMA 3R
enclosure.
The box was designed for 48 Vdc nominal systems (4 modules in series) not 60
Vdc systems(5 modules in series).

I don't know why the box failed since I was not there to witness it but I
can propose some possibilities.
This doesn't fix the problem but hopefully can explain why it happened.

>>>>

Given that the failure occurred during the fall and not initially during the
summer I initially suspected that the failure was temperature related.
I normally use Voc = 22 Vdc for back of envelope temperature calculations
for crystalline.

Since you were using DV's and have the actual Voc we can use the 21.2 Vdc
for the Voc.
Voc = 21.2 per module * 5 = 106 Vdc.
Now, 125/106 = 1.179 and from the temperature derates in NEC Table
690-7,this corresponds to an ambient temperature of 13 to -4 degrees F.
For reference, the derate for 31 to 14 Degrees F is 1.14
I don't recall where this was installed but it seems like we are in the
ballpark for an over voltage failure related to temperature.
Given the data, I still think an arc caused by over voltage is a plausible
explanation.

For contrast, using my back of the envelope calculation for a four module
system would yield Voc = 22 * 4 = 88
The NEC correction factor for -5 to -40 F gives us a correction factor of
1.25 => 88* 1.25 = 110 which is still below the product's maximum voltage
rating of 125 Vdc.
This is why I designed the product for only 4 modules in series with a 125
Vdc maximum.

While I am no longer at Pulse, I did do the design on the TCB-10 / PCB-10
combiner and I do not think the failure was caused by a manufacturing
defect.  Instead I think it illustrates the accuracy of the UL spacing
requirements and the need for the temperature derating tables in the NEC.

>>>>

P.S.

While it may be a bit of a stretch, one other possibility is insect
infestation.
The combiner could look like a nice warm home to a cold bug.
A bug crawls inside across the PCB traces and zap!
The arc forms and destroys the evidence of the bug and then continues arcing
until all the fuel is gone.
I have seen some amazing damage caused by arcing bugs down in Costa Rica.
This is one of the reasons for sealed enclosures and or screened vents in
enclosures.

Best Regards,

John Berdner

-----Original Message-----
From: Travis Creswell, Ozark Solar [mailto:ozsolar at ipa.net]
Sent: Monday, March 12, 2001 8:42 AM
To: RE-wrenches at topica.com
Subject: TCB10 Failure was (QO Breakers and Panels) [RE-wrenches]

Bill has a point here. And I have was I suspect is evidence but the doesn't
release Pulse from what appears to be a manufacturing defect.  They or Trace
still owes me a new combiner box!  Maybe I applied it wrong and I can accept
that but hopefully all of us can learn from it.

This summer I reworked a battery less water pumping system for the USFS.
The array consisted of twenty 50 watt DuraVolts which basically are Kyocera
LA51's with a stainless steel back and a clear tedlar front instead of
glass.  There was 4 sets of 5 in series.  Power point voltage is right at 85
volts but the system almost always runs at 75 volts due the AY McDonald pump
controller.  OCV @ 77F is 21.2x5=106 volts, SSC is < than 13 amps.  This is
from various spec sheets that I have found and some of you sent me.  On a
side note it turned out we were dealing with 2nd generation models and they
did not have the high failure rate (100%) of the originals.

Back to the failure part of the story.  I didn't see it happen but the
combiner box that I added caught on fire.  The system had no combiner box
lightning protection before that point. The fire appears to have started on
the back and last for a while.  It was still sparking/smoking when the USFS
employee checked on it.  He had to cut the wires in the pull boxes near the
base of the trackers to stop it.  No signs of lightning damage anywhere.
The LA-302-DC is shows no external damage nor does the TVSS (ZNR's) built
into the terminal board of the TCB10.

This device is rated for 125 volts DC.  I installed the TCB10 combiner box
in mid summer and it failed by  early fall.  From that it can be assumed
that it never got anywhere near cold enough to get the OCV of the array up
to 125v.  I emailed JPG.s to Allen Lowe of Trace and he had no idea what
happened and suggested maybe a loose wire.  I don't think that was it as the
fire started on the back of the board.

A Pulse rep. who no longer works there but was the only person from Pulse I
could get with.  He is on this list and I let him identify himself if he
wants to but you all probably already figured out.  His reply was that the
unit was designed for 4 modules in series not 5 and so he felt the over
voltage got it.  But I don't think so as I don't see how the volts got
anywhere near 125.

And please feel free to tell me I misapplied it along with the definition do
we can all learn from my mistake.

Travis Creswell
Ozark Solar
----------
>From: Bill Brooks <billbrooks7 at earthlink.net>
>To: RE-wrenches at topica.com
>Subject: RE: QO Breakers and Panels [RE-wrenches]
>Date: Tue, Mar 6, 2001, 10:43 PM
>

 One of the real hazards we have is
>counter-intuitive--since we have a limited current source (PV array) it
does
>not tend to fuse wires and do things as are cited below. It just burns and
>burns until there is not longer any fuel. It can take a long time to use up
>all the fuel in many systems were we have plastics and EVA and wire
>insulation, etc....

- - - - - - -
To send a message:
 RE-wrenches at topica.com

The archive of previous messages:
 http://www.topica.com/lists/RE-wrenches/

List rules & etiquette:
 http://www.mrsharkey.com/wrenches/etiquete.htm

To unsubscribe send a message to:
 RE-wrenches-unsubscribe at topica.com

To check out the other RE-Wrench participants:
 www.mrsharkey.com/wrenches/index.html

Hosted by Home Power magazine:
 www.homepower.com

For info contact list moderator by email:
 michael.welch at homepower.com

____________________________________________________________
T O P I C A  -- Learn More. Surf Less.
Newsletters, Tips and Discussions on Topics You Choose.
http://www.topica.com/partner/tag01

- - - - - - -
To send a message:
 RE-wrenches at topica.com

The archive of previous messages: 
 http://www.topica.com/lists/RE-wrenches/

List rules & etiquette:
 http://www.mrsharkey.com/wrenches/etiquete.htm

To unsubscribe send a message to: 
 RE-wrenches-unsubscribe at topica.com

To check out the other RE-Wrench participants:
 www.mrsharkey.com/wrenches/index.html

Hosted by Home Power magazine: 
 www.homepower.com

For info contact list moderator by email:
 michael.welch at homepower.com

____________________________________________________________
T O P I C A  -- Learn More. Surf Less. 
Newsletters, Tips and Discussions on Topics You Choose.
http://www.topica.com/partner/tag01