Re: [TCML] MMC Caps

[Gary Lau <glau1024@xxxxxxxxx>]

The data sheet for the 940C series says:
Type 940C round, axial leaded film capacitors have polypropylene film
and dual metallized electrodes for both self healing properties and high
peak current carrying capability (dV/dt). This series features low ESR
characteristics, excellent high frequency and high voltage capabilities.

For the 942C series:
Type 942C round, axial film capacitors utilize a hybrid section design of
polypropylene film, metal foils and metallized polypropylene dielectric to
achieve
both high peak current as well as superior rms current ratings. This series
is ideal
for high pulse operation and high peak current circuits.

The self-healing is accomplished by having a floating plate between the end
plates, made of a very thin metalized layer.  Should the dielectric between
an end plate and the floating plate break down, the thin metalized layer
vaporizes in the small area of the defect and leaves no electrical path
between the end plates.  The loss of the small area of floating plate is
inconsequential.

Odd that only the 940C description touts the self-healing feature, but I
believe that both types are capable of this,as both types employ the
floating plate, to achieve both a higher AC voltage rating (doubles the
corona inception voltage), as well as the self healing trick.  The
significant difference between the two types is that the end plates are foil
on the 942C's and metalized polyester (?) on the 940C's.

Regards, Gary Lau
MA, USA

On Wed, Aug 11, 2010 at 10:59 AM, jimlux <jimlux@xxxxxxxxxxxxx> wrote:

> bartb wrote:
>
>> Hi All,
>>
>> Before assuming too much, maybe dig into the TCML archives to find out the
>> history of MMC's in TC use. The "guy" referred to below is none
>>
>
>
> Indeed, Bart's advice is good..
>
> As I recall, the 940 and 942 have superficially similar ratings (942 is a
> bit better) as did the WIMA MKP, I think, but the difference was in the
> internal construction (Terry and others cut a whole raft of commercial
> capacitors open after testing to do Destructive Parts Analysis)
>
> The difference had to do with the partial failure..the 942s are basically a
> whole bunch of smaller caps in parallel internally, each individually fused
> to the external connection.  The 942s would punch through in one segment,
> the internal fuse to that segment would blow, but the overall cap would
> remain good, just slightly lower capacitance.  The 940s didn't have this
> type of construction, so a overvoltage that punched through killed the cap.
>  (I may be misremembering here...)
>
> But overall, the idea was that the 942, in addition to having better
> ratings, was MUCH more failure tolerant.  This allowed running much closer
> to the edge (if not well beyond it) on the ratings (e.g. running a 1000WVDC
> cap on 1000VAC rms, which is actually 1400V peak).
>
>
> There were also some thermal analyses done.. the 940 has higher ESR than
> the 942, so requires more derating when run with high peak currents.
>
> One can certainly design a TC using the databook values for the caps, using
> suitable derating, etc.  But most hobby TC builders are cost sensitive, and
> want to push components as much as possible (e.g. spend $30 on 10 942s at
> $3, at the ragged edge, rather than $60 on 30 940s at $2 each, running at
> databook values).
>
> The capacitor count adds up quickly when you start stringing more in series
> to get voltage margin, and then you have to string more in parallel to get
> the capacitance back up (and increase the current rating)...
>
> I think that as of when the analysis was first done, the economics said
> 942s are a "better deal" than the 940s
>
>
>
>
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxx
> http://www.pupman.com/mailman/listinfo/tesla
>
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla