Mail Archives: geda-user/2023/05/10/17:27:26
Hi Bob, and Frank and Gabriel,
Thank you for your insights.
This part of the application is very simple: it's a linear audio
compression filter with a long release time constant (some 45 seconds),
with space constraints making smallish timing capacitors desirable. Foil
capacitors are used for reasons of low leakage current and good
linearity, but these quickly become quite bulky with capacitance
exceeding 1 uF. Most MLCCs with X-type dielectric have serious
capacitance derating with increased voltage (15 volts maximum), so these
were avoided.
However, from your answers I gather that the chosen solution with 22 M
resistors is also not optimal, and that potting or coating does not
guarantee trouble-free long-term operation.
Maybe the best solution here is to use high-capacitance MLCCs after all
with a maximum 1 M timing resistor, but only use a tiny charging voltage
(e.g. 0.1 volts), and amplify this voltage with an extra opamp.
Best regards,
Richard
Op 10-05-2023 om 21:17 schreef Bob Paddock (graceindustries AT gmail DOT com)
[via geda-user AT delorie DOT com]:
> On Wed, May 10, 2023 at 7:45 AM Richard Rasker (rasker AT linetec DOT nl)
> [via geda-user AT delorie DOT com] <geda-user AT delorie DOT com> wrote:
>
>> ... 22 Meg ...
>> Would it be better to use a larger resistor (e.g. 1206), or forego SMD
>> altogether? I can of course also specify a moisture-resistant PCB
>> coating, but that is a rather expensive extra production step.
> Coatings are not hermetic seals.
>
> They are good at keeping out particle contamination.
> However moisture will still seep through over time.
>
> As it is now devoid of contamination it will never be noticeable in a
> digital circuit, due to its high impedance;
> perfectly pure water is an isolator.
>
> However in high impedances, or RF tuning circuits, the dielectric changes.
> Perfectly tuned transmitters are returned in six months for being off frequency.
> Been there, done that. :-( Water ran off the boards in removing the coating.
>
> This is why some high impedance related parts. such as Electrostatic
> Grade Op-Amps, say "expose to air" in their data sheets.
>
> As Frank pointed out, flux is an issue.
>
> We really try hard to avoid QFNs and anything that is hard to clean
> under in our low current (uAs) standby products,
> Any residual flux will build up contamination over time, which could
> be measured in months or years.
> A 10 uA poorly cleaned board built and measured today, might measure
> 100 uA in six months.
> Been there, done that too.
>
> If there are no other through-hole parts consider putting two 10 Megs in series.
> To avoid the labor costs of the stuffing charges.
>
> Knowing more about the overall circuit may lead to better/different answers.
>
Met vriendelijke groeten,
Richard Rasker
--
Linetec
Akkerstafhof 15
7544SP Enschede
053-4350834
http://www.linetec.nl/
- Raw text -