[Andrew Holme]

I'm developing a frequency synthesizer.

I was worried about ceramic capacitor microphonics, so I used these film
capacitors in my loop filter:

Panasonic ECPU1C105MA5 1uF, 16V, 1210, PPS (Farnell 969-4315)
AVX CB017D0223JBA 22 NF, 50V, 1206, PEN (Farnell 165-7961)

The 1uF is microphonic. Tapping it with a plastic rod produces FM on the
VCO equivalent to ~ 20uV amplitude on the control voltage.

Looks like I'll have to try a ceramic or a tantalum. Leakage shouldn't be a
problem as I'm using an AD9901-style PFD which behaves like an XOR gate PD
when locked.

Any comments or advice?

TIA
Andrew.

 

[Jim Thompson]

Don't tap it with a plastic rod?

John

Sno-o-o-ort ;-)

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |

I love to cook with wine Sometimes I even put it in the food

 

[Spehro Pefhany]

Ceramics can be FAR WORSE...
If you can find a "film" capacitor where the plastic sheet has been
coated by sputtering and/or depositon, that constuctionshould be the
least sensitive.

Has anyone compared through-hole film caps vs. SMT as far as
microphonics goes?

 

[Spehro Pefhany]

I'm developing a frequency synthesizer.

I was worried about ceramic capacitor microphonics, so I used these film
capacitors in my loop filter:

Panasonic ECPU1C105MA5 1uF, 16V, 1210, PPS (Farnell 969-4315)
AVX CB017D0223JBA 22 NF, 50V, 1206, PEN (Farnell 165-7961)

The 1uF is microphonic. Tapping it with a plastic rod produces FM on the
VCO equivalent to ~ 20uV amplitude on the control voltage.

Looks like I'll have to try a ceramic or a tantalum. Leakage shouldn't be a
problem as I'm using an AD9901-style PFD which behaves like an XOR gate PD
when locked.

Any comments or advice?

If possible, avoid having any substantial DC voltage across the cap--
that will minimize voltage fluctuations as a result of transient
changes in capacitance.

 

[Andrew Holme]

If possible, avoid having any substantial DC voltage across the cap--
that will minimize voltage fluctuations as a result of transient
changes in capacitance.

Just tried this. At the bottom end of my tuning range, where the cap
voltage is small, I see no detectable microphonics. At the top end, where I
have almost 15V on the cap, hitting the board produces a large frequency
deviation spike. I'm very dissappointed with this Panasonic film cap; it
looks nice, with rainbow colours as it catches the light; but it doesn't
work.

 

[Baron]

Just tried this. At the bottom end of my tuning range, where the cap
voltage is small, I see no detectable microphonics. At the top end,
where I have almost 15V on the cap, hitting the board produces a large
frequency deviation spike. I'm very dissappointed with this Panasonic
film cap; it looks nice, with rainbow colours as it catches the light;
but it doesn't work.

Dow Corning do some nice glass caps but they may not have the values you
want.

 

[Rich Grise]

Don't tap it with a plastic rod?

LOL!

("Doctor, it hurts when I do this."
"So, don't do that.")

Cheers!
Rich

 

[Andrew Holme]

I'm developing a frequency synthesizer.

I was worried about ceramic capacitor microphonics, so I used these film
capacitors in my loop filter:

Panasonic ECPU1C105MA5 1uF, 16V, 1210, PPS (Farnell 969-4315)
AVX CB017D0223JBA 22 NF, 50V, 1206, PEN (Farnell 165-7961)

The 1uF is microphonic. Tapping it with a plastic rod produces FM on the
VCO equivalent to ~ 20uV amplitude on the control voltage.

I've now tried:

1. Film 1210 16V (the original)
2. Tantalum 0603 16V
3. Ceramic X7R 1210 50V
4. Radial aluminium electrolytic 63V (15 years old)

Guess which:
a. was better than the film but still a bit microphonic.
b. can hear a pin drop next door.
c. has no microphonics whatsoever.

 

[Andrew Holme]

Betcha the X7R was the worst, with the Olde Sprague being the best...

Correct

 

[Baron]

Are they still being made??

As far as I know ! Someone posted a link to the manufacturers data
sheets recently. I belive it was AVX. I have a copy of the pdf if you
want it.

 

[Eeyore]

I'm developing a frequency synthesizer.

I was worried about ceramic capacitor microphonics, so I used these film
capacitors in my loop filter:

Panasonic ECPU1C105MA5 1uF, 16V, 1210, PPS (Farnell 969-4315)
AVX CB017D0223JBA 22 NF, 50V, 1206, PEN (Farnell 165-7961)

The 1uF is microphonic. Tapping it with a plastic rod produces FM on the
VCO equivalent to ~ 20uV amplitude on the control voltage.

Looks like I'll have to try a ceramic or a tantalum. Leakage shouldn't be a
problem as I'm using an AD9901-style PFD which behaves like an XOR gate PD
when locked.

Any comments or advice?

SMT caps have a dreadful reputation in audio. Try substituting with a
polypropylene or even polyester type with leads temporarily.

SMT resistors can be pretty crappy too.

Graham

 

[Eeyore]

Has anyone compared through-hole film caps vs. SMT as far as
microphonics goes?

Serious audio uses through hole components for all passives. WE KNOW there's a
difference.

Graham

 

[Eeyore]

Thanks for the tip. Unfortunately, these are loop filter integrator
caps, and need to have between 1 and 15V across them.

The previous version of my board had through hole filter components
(Poly Layer caps) and that is microphonic also I discovered last
night.

My options seem to be:
1. Tantalums
2. Re-design filter with smaller caps (could try NPO/COG) and larger
resistors
3. Miniature aluminium electrolytics

Aluminium electrolytics can be the most microphonic of all.

Graham

 

[Eeyore]

Aluminum electrolytic should be fine for your low-pass application.
That dielectric, Al2O3, is definitely NOT piezoelectric.

The most microphonic caps I've ever met were poorly manufactured electrolytics.
The 'capacitor pill' wasn't securely located in the can. The re-work that involved
! Thanks purchasing.

Then there was the batch of TL072s the MD insisted on buying cheap ( they'd been
soaked in a leaky container during shipping ). LMAO. What the guys behind desks do
is something else.

Graham

 

[Eeyore]

Funny thing though is surface mount resistors measure as about the
most ideal resistor you could imagine. The ESL is just the electrical
length.

Sadly the surface mount resistive film has more excess noise than metal film leaded. You
can HEAR the difference. The material used is grossly inferior. It's even on the data
sheets where the manufacturer dains to publish the numbers.

Plus, the dielectric films used in SMT caps and the construction don't match what we
audio pros are used to.

Graham


due to the hugely increased level of spam please make the obvious adjustment to my email
address

 

[Eeyore]

Generally the guys behind the desk sink the company, but make sure
they have a golden parachute.

Time for change. Politics and dumb politicians are no different. I LOATHE most of them.
There a FEW good 'uns though, bless them.

Graham