On Tue, 15 Sep 1998, Morgan Olsson wrote:

> >5 m/s^2 ~= +/- 0.5 g ?
> 
> Right.

Ok, you're in range for piezos, these do 2 g's.

> >but what do you mean by cheap,
> About 30$ incl glue cirquit...

The piezos are much cheaper than that. 

> >and do they have to read accurately in static conditions ? 
> No.
> I will use it to stabilize mechanic oscillations in the 0,5 to 2 Hz range.

0.5 Hz is tight. See below:

> >There are piezo
> >accelerometers that do not read out static acceleration, only delta (e.g.
> >Murata, Tokin etc). These can be used by configuring the system as an
> >integrator. They drift some so it needs to have a deadband and a zeroing
> >facility.
> 
> Souds interesting!

Wait until you have one working. This is a really cool way to measure
delta-A, and low power too.
 
> Have you got more specifiv part of part no or URL i can test tomorrow?

Murata: PKGA-S60A (I've just been there, full data sheet on the web, but
low limit 0.7 Hz. It will work at 0.5 but you have to derate sensitivity).

Go to Murata category search: you have Acceleration sensors, gyrostars
(you don't want this for your app) and shock sensors that are of interest
for your app imho. 
 
If this is an automotive app, you'll need good low pass filtering imho.

> Althuogh ADXL seem best as i will only ship 50 units per year and don+t
> want too much design hassle... or maybe it is quite simple as i donīt need
> static operation.

mmh. Murata have minimum shipping quantities. Try the Swedish Murata
dealer for a broken reel or returns...

Peter