[sdiy] There is no fun analogue chips anymore!

[Gordon JC Pearce]

On 28/06/12 21:19, Barry Klein wrote:
> Many designers don't realize that effective capacitance drops often 50% with DC across multilayer type ceramics - and this varies between manufacturers and voltage ratings.
> I hadn't come across any that got worse over time though.  Any background on this - a particular vendor?
>
> Barry

I've no idea how to tell what the particular vendor of the chip is. 
What I find is that some equipment has stock faults that can be traced 
to ceramic caps going leaky over time.  The Kenwood mic preamps are one, 
and another example is the RF amp and squelch buffer in Tait T2000 
Series 2 radios.

In the Kenwood, they go leaky and the DC conditions around the mic 
preamp change giving crappy distorted audio.  In the Taits, much the 
same thing except the DC bias conditions around one of the RF amps 
changes (or is it an IF amp?  Need to check!) and they go progressively 
deaf.  Normally as you reduce the input signal, you get smooth noise 
around -100dBm and a sputtery quality at around -118dBm.  When the RF 
amp fails, it starts off with the sputtery 12dB SINAD point about 
-110dBm and gradually gets worse until it's around -90dBm by which point 
the radio is clearly sick!

Oddly enough, these are all surface-mount capacitors, but I've had disc 
ceramics perform similar stunts before.  Now, someone mentioned to me 
recently that they had a crystal filter fail because of a DC bias across 
it causing the silver plating to migrate across the crystal and short 
sections out - could the same happen inside disc ceramics?

-- 
Gordonjcp MM0YEQ